What went wrong with BeOS? Why did it fail?

What went wrong with BeOS? Why did it fail?

Attached: be-os.png (1600x1600, 31K)

Other urls found in this thread:

osnews.com/story/29536/Apple_engineer_tried_to_buy_BeOS_from_Palm_for_the_iPhone
en.wikipedia.org/wiki/RISC_OS
twitter.com/SFWRedditVideos

Microsoft fucked them.

Commercial reasons, not technical ones.
>Only ran on niche architectures.
>Until it was far too late.

expensive proprietary hardware platform
limited market presence
niche focus
advantages over dominant platforms consumers didn't care about

>didn't care about
They didn't know to care about.
The OS was better than Windows (and unfortunately still is).
The system kits, very clever. Very amiga-aware, with awesome features like datatypes.
System servers, thus native SMP rather than a patch approach with a monolithic kernel+locks.
Very low scheduler latency. Very good at running realtime pipelines, such as music production ones.
But... only ran on niche platform they failed to promote properly.
Meantime, Microsoft made a sweet little monopoly, and past that point the majority of the world has been stuck with shit and nobody's managed to change things for the better.

>muh BeOS
There are far, far better examples of much better operating systems that failed in the end, yet Jow Forums mostly talks about BeOS/Haiku.

Also, what a dumb question.

>The system kits, very clever. Very amiga-aware, with awesome features like datatypes.
who cares
>System servers, thus native SMP rather than a patch approach with a monolithic kernel+locks.
who cares
>Very low scheduler latency. Very good at running realtime pipelines, such as music production ones.
who cares beyond a niche that was already well serviced by existing platforms

people who are getting a job done rather than embracing a lifestyle don't care if a platform appeals to you as a developer or works some nerds up into a frothy lather, all they see is an expensive as shit proprietary box with terrible support, an uncertain future and less overall versatility, who-cares details like the fucking scheduler are absolutely secondary to these concerns

this isn't to say that I dislike BeOS or consider even NT to be architecturally superior in every way, getting a BeBox is on my bucket list, but I'm not going to lament about a future that would never exist or whine about why they didn't take over the world, like that was even their intention

>all they see is an expensive as shit proprietary box with terrible support, an uncertain future and less overall versatility,
No, they don't. They simply never heard about BeOS, much less its advantages.
But I'm with you in the "don't care". It was just another company pushing a proprietary OS.

no, you're definitely not at all wrong there, just saying that assuming a given person has heard of it and is weighing it with other platforms, that would definitely be some concerns they'd have if they really went in-depth on it

Apple didn't buy it. They bought NeXT instead.
They didn't go for plan Be. Gassee wanted too much and overplayed his hand.
What would Be be like if Apple had bought it?

>It was just another company pushing a proprietary OS.
Just like Microsoft, and we know what happened there.

Applel engineers who had help develop it wanted to buy it when applel was developing the iphone.

osnews.com/story/29536/Apple_engineer_tried_to_buy_BeOS_from_Palm_for_the_iPhone

It's not like Palm ever did anything with it.

>There are far, far better examples of much better operating systems that failed in the end
care to name some? Im honestly curious, because other than Amiga I dont think there were hardly any other graphical/multithreaded OS's made

Which is weird that they didn't take the 800k

Unix workstations, Lisp machines, Alto derivatives? There were fuck tons of graphical/SMP-aware systems in the high end since the early '80s.

Now this is a thread that's Jow Forums-worthy.

Actually Jow Forums seems to obsess way more over Plan 9 or TempleOS.

NeXT, arguably, though that got rolled into OS X. Plan 9 from Bell Labs. OS/2.

Later BeOS wasn't high-end though.

And everyone who tried the PE was blown away by how much faster and more stable it was compared to their Windows installation on the same computer.

He didn't specify a specific market.
>And everyone who tried the PE was blown away by how much faster and more stable it was compared to their Windows installation on the same computer.
You'd probably get the same reaction from them at NT too, dog shit was more stable than 9x.

You didn't get the speed though.

BeOS was absolutely amazing in that regard. It booted up in a few seconds, the interface never lagged, it could play multiple videos simultaneously without a problem on the same machine that struggled to play one under Windows.

the marketplace decided they were inferior to Microsoft. praise the free market.
it's better to use what other people use than 'technically' superior crap no one else uses.

I was there.

Bunch of assholes ran the company.

It was not that great anyway.

So. Asshole to talent/performance ratio was off.

Attached: B9988E82-D542-4BAA-B1D9-F9C369F0EB03.jpg (768x984, 103K)

>PE
Tried it around 2001. It was indeed much better than win2k and Linux.

Lisp machines weren't dying by the 80s and unix workstations are still identical to the way they were in the mid-80s/90s.

it was good

low-quality bait

BeOS is a great example of how being being a superior product doesn't mean you'll win
just as Apple products are great examples of how being technically inferior doesn't mean you'll lose

Yeah it wasn't like Windows came pre-installed on everything. People all just picked it. Sure.

The free market said people wanted Windows because they bought the PCs with Windows pre-installed.

I'm pretty confused as to what this has to do with anything I've posted or the post I quoted.
He seemed to be under the illusion that Commodore and Be were the only companies that ever built multithreading/graphical operating systems for some reason, which is absolutely false.

Well aside from the stuff that isn't dead or Commodore/Be, what's left? OS/2?

I know a few but...
UNIX -> we all know that one
RiscOS -> coop multitasking aka shit
Apple -> coop multitasking aka shit

unix is still alive, riscos was just unix for risc

This -- If Gassee didn't get greedy, it would have been the first choice. It didn't do much at the time, and would have taken a lot of work to get ready. Along comes jobs with a functional NeXT, and he's willing to hand it over for a price.

While BeOS could have been a superior OS in the end, it came down to the bottom line. Always does.

>riscos was just unix for risc
Facedesk. Not even an attempt to educate oneself before speaking with an authoritative tone on a subject you have no clue about.
en.wikipedia.org/wiki/RISC_OS

This cancer is why Jow Forums is shit these days.

>dead
Maybe Xerox and Lisp machines by the end of the '80s, but System V vendors were huge, dude.
But yeah, I was just thinking as I wrote that post that I forgot about OS/2 as well. Plus god knows what other one-off unique platforms and attempts were made that nobody cares to remember off the top of their heads.

Actually in the 90's you had to visit your local office computer ore computer store and buy a copy of windows or use a friends. You actually had to go out and get the software and leave your cave.

I meant that unix/xorg and vaguely macOS are alive and kicking in 2018.

Nah even 3.1/3.11 were OEM'd.

I get what you mean now.

>Nah even 3.1/3.11 were OEM'd.
Sometimes it would be OEM depending on where you bought your computer. Updates and software in general had to be purchased manually or given to you by friends or coworkers. That's why terms like sneakernet existed. Assuming all machines had windows pre-installed is disingenuous. Even if we assume all had Windows 3 or 3.1/3.11 etc installed you still had to manually purchase or acquire a new version of windows to install.

Well, software in general was varied. Plenty of people had dialup internet, or even BBSes, for software updates before win95. Of course if you DID buy a win3.1 to 95 UPGRADE it was like 400 floppy disks without a cdrom. CDROM was the really rare early.

>400 floppy disks
Not that many.

Yeah that was hyperbole. It came in like a long sleeve though, it was fucking awful.

I was still with Amiga back then, but ran Linux on my first PCs and it was awesome. Windows 95/98 was such trash.

Yep, 95 and 98 were super crashy. I eventually had an NT4 machine that could run most of the windows stuff without imploding though. Had a bitchin irix workstation I couldn't do anything with and got into 90s linux. Nowadays I don't have any windows.

Still want haiku to succeed because it is qt.

They got greedy when Apple tried to buy them

- microsoft licensing restrictions against selling dual-boot systems
- nowhere near enough killer applications
- the BeBox was actually not as good at multitasking as it should have been for the price (aka PPC 603 was a mistake)

>(aka PPC 603 was a mistake)
PPC 603 is about as good as CPUs got back in the day. Affordable alternatives included CPUs based on ARM and x86. Both were, back then, significantly slower than a ppc 603.
>The BeBox was a mistake
There we agree. They should have focused on makign the software run well with the popular existing platforms, rather than try to tie it into a hardware platform.
When they finally targeted x86, it was too late.

>There we agree. They should have focused on makign the software run well with the popular existing platforms, rather than try to tie it into a hardware platform.
one of the major advantages it had was SMP before your common PC had more than one cpu/core. it still did a better job at multitasking on a single cpu than windows did from what i've read

>PPC 603 is about as good as CPUs got back in the day. Affordable alternatives included CPUs based on ARM and x86. Both were, back then, significantly slower than a ppc 603.
How far up your ass did you reach to pull this out? The 603 was the low-end of the second-gen line and clock-for-clock slower than even the 601. It was pretty much Pentium equivalent, the P6-equivalent 604 was the one worth having, even though it wasn't all that stellar either and ultimately just a low-end workstation chip.

ARM sure as fuck wasn't an alternative either, aside from propping up the dying Acorn platform it was all but dead on the desktop by then and had shifted into embedded systems.

>It was pretty much Pentium equivalent
Even a 68060 was ~2x the IPC of pentium... I'm confused.

>Along comes jobs with a functional NeXT, and he's willing to hand it over for a price.
Jobs asked for more money than Gassee did!

Wasn't the BeBox a dual CPU system?

it's not a unix.

Their business model was to be bought by Apple to replace Mac OS Classic. However, Apple bought NextStep for that instead (as they came with Steve Jobs)

BeOS came with Gassee, who was more popular among Apple employees than Jobs.

>I'm confused.
Me too. What exactly does the nebulous and ill-defined "IPC" of the completely unrelated and utterly forgotten 68060 have to do with the overall performance of the radically different PowerPC 603 when compared to contemporary Pentium processors?

Even Motorola themselves didn't pretend it was the fastest thing around, the 603 was a cut-down and low-power chip from the beginning that mostly saw its way into mobiles and low-end Macs.

>but he hurt some nerd fee-fees!
Nobody cares, the point was saving the company, and round two fifty-five or die wasn't going to do it.